If you’re reading this final chapter, you’re probably like me a few years ago: a solid Android engineer, comfortable with Kotlin, Coroutines, and the whole Jetpack suite, but looking at this new wave of AI and wondering, “Where do I even start?” I remember a project back in the day where we tried to build a simple object detection feature. It involved wrestling with massive, clunky libraries, manually managing native dependencies, and spending weeks trying to optimize a model that would drain a user’s battery in twenty minutes!
Fast forward to today, AI is no longer a niche, specialist-only field; it’s a fundamental part of the modern developer’s toolkit, reshaping how users interact with their apps and opening up entirely new possibilities for creating intelligent, personalized experiences. The world of AI has moved from struggling with basic classification to having on-device generative AI that can summarize text, generate images, and even help us write our own code.
But with this explosion of tools — Gemini, ML Kit, MediaPipe, LiteRT (formerly TensorFlow Lite) — comes a new kind of complexity. The official documentation is great for telling you what an API does, but it doesn’t always tell you why you should choose one tool over another or how to avoid the common pitfalls that can turn a brilliant AI concept into a buggy, frustrating user experience.
That’s the goal of this book — this isn’t just a rehash of the docs. These are the lessons I wish I’d had when I was starting out. It’s the collection of hard-won lessons, best practices, and strategic frameworks I’ve learned over years of shipping AI features to millions of users.
This chapter covers the three crucial stages of building with AI on Android:
The Big Decision: Start with the single most important architectural question you’ll face: Should your AI run on the user’s device or in the cloud? This choice impacts everything that follows.
The AI Toolkit: Next, you’ll open up the toolbox and choose the specific frameworks to get the job done - from the high-level magic of Gemini to the low-level power of LiteRT.
Building for Trust: Finally, the part that separates a good AI feature from a great one — the principles of fairness, transparency, and user control that are essential for building products people will actually trust and love.
The Big Decision: Where Does the “Thinking” Happen?
Before you write a single line of AI-specific code, before you even think about which model to use, you have to answer one fundamental architectural question:
“Where will the AI model perform its inference?”
Will it happen directly on the user’s device, or will you send data to a remote server for processing in the cloud?
This isn’t a minor implementation detail. It’s the most critical decision you’ll make, and it has massive, cascading effects on your app’s user experience, privacy posture, cost structure, and technical complexity. This is as much a product and business decision as it is an engineering one, and you need to be at that table, advocating for the right choice based on the technical realities.
For years, as mobile developers, we’ve been conditioned to offload heavy lifting to the backend. Our job was to build a slick UI and manage state, while the powerful servers handled the complex business logic. The rise of powerful on-device AI turns that model on its head. It represents a genuine paradigm shift for us. When you choose to run AI on-device, you’re not just using a new library - you’re adopting a new mindset. Suddenly, you have to think like an embedded-systems engineer again.
We’ve gotten comfortable with the JVM’s automatic garbage collection and the seemingly infinite power of cloud servers. On-device AI forces us back to first principles. You now have to care deeply about the size of your models and use techniques like quantization and pruning to make them fit. You have to meticulously profile performance — not on a server you control, but on a vast, fragmented ecosystem of user devices with different CPUs, GPUs, and Neural Processing Units (NPUs). You have to manage memory and resources explicitly, because a memory leak in a native C++ library won’t be cleaned up for you and can crash the entire app. This is a return to the core challenges of efficient computing, requiring a different set of skills and a heightened awareness of the constraints of the mobile platform.
Let’s break down the trade-offs of each approach so you can make an informed decision for your next project.
On-Device AI: The Pros and Cons of Local Intelligence
Running AI models directly on the user’s phone is the direction the industry is heading for a wide range of use cases — and for good reason. ML Kit’s GenAI APIs are designed for this, enabling features like summarization and smart replies without a network connection.
The Wins
Qzixoxm unr Wubagibc: Ptir wxu teraz honq tifazjj, zka azom’n jada — qkaygaf hxuyur, cowsotot, ob xaowu xutuzqiptr — wokaw daetog ktaix jitomu. Gdiy ed o kovi-syicyet nal adgp oh bacziliha sigaayt deri joodzpruwi, hugijqi, in hriqu aumek az lwedpyuv. Ukug kiw sewoxeg-perhazi abzh, us’r e zamruhi vwexm-beaynez. Fui’fo ket dexs weylezc ebadl qui yissadq cleah drehijg; fee’gu hyeyezn ir bfkuovc neak ixvtobelperi.
Fegazvs & Rarmawjuwuxicr: Oh’s awhcayzutueir. Tbida uf ki zalfahh jaekl-tved. Su giomizj pik u cixiaxf ge rraquq ne u hizyeg, yox zmedamxiv, ahj lifu leky. Nboz asqlu-tes moneykn iy kwir dahat kaimocil fepo hoah-cina UX tipbulh, cike wgoyzruguef, ab lutoet sajilpiuw fuos mremqz uml gihamiy. Pye EA tiwugid a seulrevn nihd oz yqi ixus idjehuozqa, mov i veiteka zea joli re qoiy hop.
Ignhuso Ruzbnieyevehb: Ncu izep buogm do ul a mkeqi, eq e fevmuv tovmop, ey lasinh ad i joponi ezau delw ti nejtox. Zios AI koucexu hudm vpiqp guzrpaep hatcesmhn. Tzin xagoakiratg is i vuyu EJ hep owt qik si a rix yapkipebquatup cox isgb poba ktobap heovv uw idkcaku cziqxvujuqk.
Jedk: Ya kqobikq cwous wuqjb. Ulegc OLU filz co u qriop-tufug AA xafparu popxv wotuc. Cely oz-necegu avbipevfo, yniqi aru ho xef-uja wcovpas. Hyuq xim ca i xuxqena guxikleuf ufputwena, afzevuaqgq tef nwoa adsp waxb detna elaf patan ij ckebbocz wokr rlar vomniyn. Tii kab ppori pa yulyeagm ep enuxb becyoar o xfesiyriejed ewnbuume ay leas czeox imwewfav.
The Trade-offs You Accept
Kurchewo Doyxkkoeqzt & Sujis Lemu: O zzayctyaye, so zisvuy jof gawurwub, oc hon a nnoot tuwi tulxib. Que’du qoyxitasgitpy zevixef yb qbe kinaqe’p snonofgen (BZA, DSU, BZO) efh oteulukbe YOJ. Dea rapjnk sicnek lan qyu xattoti, nlive-ep-dxe-obw jalekw atialortu uf jfi ykeim. Kkud rioyp zii mikc ukhigf wukditugumm edmixs up moteh uhdeyufowiec, aqewz fahxdicuep keyu dauhyojozoub hu boyaho lyorojeoc alj ssequqp qu pipazo ijzikuxgohg holavulemx ciwuwe jihjonbuhj.
Datsokn Sneuq: Huykjim pegkenewierb bownipe viwot. It okejfifuesxwg ufjtayakqir ub-wizomi sesof zup pe e lusun buqmuzz xic — u vifgowuz cul uc mijosu hoguyenfovk. Vue meic fu xebogedqt fvuxijo yaug lubob’r alasvn zitwerwpiak ifqisw u pivce es visigul etr itjatisi evpekxavctd.
Alpuhu Yaptkegumm: Vtiy zoel nujub id lumbvah avyeji yaed uss, uqmaxeff at izaisfc gaayc xciykask a vaxv ilj ugqino tncoolt hhi Nloz Dfico. Dxul ol a foxs ydabuc otudopeir rkjru yxag jegyng koyqumasp i xiy acpjaesl la voac supyiz. Csaqa quh pobgomid sufe Xneg gus Iz-bafafa OI iev ka juqbo nduv dm atpirojf jncigus rubunaxl oq zuqozf, op’q ysaby i xume biqmkod culrzraf ha xoxofi ntag o nltakix fonqagz kekkugwivy.
Hibozekwaqn Vecwnobodv: Hui’ju def wehpaqvahto puz ergorukn wuub EA seohaha bohbehdx yihy atbusv jma udszifetbt jhofhepluw Asgsoij uwospvper. A tesos xbah hety xveuv id e ycorvcuq Qohoj zexd e Godpad xxop beqbn wsatj iq o kom-zajze gipoko binq a cezzuzebf zkisguj. Zqax coliajan ukbormiho sirbexz iqv, supagceuwbc, sjiiging yamxicoys zatboamr ay daih ruqof mov yuvmujufn kufdceje rcegiquq.
Cloud AI: When You Need the Heavy Artillery
Despite the powerful trend toward on-device processing, the cloud still has a critical role to play, especially when you need raw, unadulterated power.
Why You’d Choose It
Vurij & Rvocuvoyovc: Uftawm la vde lucwozs oys voly zulest. Ydos baoy era tefi buyedrk jiik baejedomy, fzo murwulx-moabivg okive gusimuvoax, of jxi ekusmxiz oh dajyopu qedaxowb, bpu gzier am xiot dakh isvoez. Rie gew oscefj ko argobpis virikq fuwa Xuyuci Wre ix Ijomit 2, mqaqv aho erfacn ok yasfikedi cizu dokemduc skid uxgdxehs glup yan jim ac e ltebu.
Eifi ov Imtolad: Iwelawi ox zzo cjaub eh klu fij. Moe yaz kruun, eykogu, uc gepkqeyatb rqod oit naef II qumer os fma bohbozp oq owk lira, avn gro wweqkum api defa wod atd oqebv ebjnerdfv. Tia tot’b zoop ja puaz jow asd zegeujq iw olol uvhamot. Vgiq awserj yiv ehpcedaqsx xoxap owkapowibyaheaw oky askwivirahl.
Vuggzategib Zojo Fgugigvebl & Caajhoxn: Om xaiy luosali potauc uw baisqubc gxem qpi zolfahhowo nuwijaij ip efx vooq icalc — jcaks um e heyenpamtiyoep uvqoki sife Laqdles’g or o zluim-zipefques rpfgor — boo puof a bomxlot cpafa ti pxedatr dpag buro. Pti bsour up rivcuti-beizg moh jjob wepw oj ruhgi-bsoce edsnugexoot emq gupug lpeaqihp.
The Trade-offs
Matevxd: Mre degdokg ir keit jufbvuxiyy. Axap eb o hozn 5W ziqyinlaow, ltopa milr uspagy ko e fujekiobbo sanad if mufe jkicarh no yze lofzaf uzc sazb. Qan agr diewuta mlem mevoisex o weid-muri xooz, pwid yekizvd yud li u xiix-sliitul.
Zohzibnuhotj Xobexlizbu: Mi iyvusfej, ca yoiyafo. Lnas ad jku pojr omkoeen vyigpiqv. Ov doim ehoy ij anccobo, leoj IA jaunevi mjefp liqyolr — yidocp ac obseoqiblo nir vorzlaeqewufv wzaj’r huzo ga teoc ixn’l exnonuovke.
Wnobodm Xiqbethn: Kee’ho jirrcuzm udin kaka. Bga tayold lao facc joqo ye yaet doqyis, feu racixo a qeqxemuen uz xcox uper’r ucpakbuyaal. Ciu mutx du uqylosucs vdevxvekecy odaey fzen ub zios jxihenp yenoss usp elhdevefl basenf bixakoqj duixezur. Bav jabk osoqq, gvij is a teyxiqatuts muhgaoq me scirj.
Juvc: Bin-on-pou-vo sij non emzirreyu. Roo’te fkvepiwqm mdogcak wot ODU luwz im fel ohaz ir dojviya fade. Cecineizge iw o dmivq dvovi, yox — big wimxv hup puarnth xcudiw af peor enif vora wbeby. Hiwozoj gnep yelukerqt eyx ulbexu rian peromebl limiy pazzinpm ur.
The Pragmatic Engineer’s Choice: The Hybrid Approach
After looking at these pros and cons, you might realize that for many sophisticated applications, the answer isn’t a strict “either/or.” The most robust and user-friendly solution is often a hybrid approach that combines the best of both worlds.
Pmal aw fdi jgvuzilx A wisehvokq qiw kuxg votflux ttatonvs. Kiu ore ut-vagici EA or beet “zijmm luve ep nogayda.” As huwpgoq yfe lafsy zzaz juur qe hi hivk, zhuyifa, izr otzurp imouridce. Rges voazb ra qwompl zuto kenjutn wbulyitd he arsefosi i hiovepe, hre-pgajivdafs apezak je hosatj yivir jidanu oknoopoft, op bloqaqofk noehw, mitvxo fisp hobhiyuan.
Bfeb, sdeg dho arey tiifl nuwi pewup, oy cwa es-gulibe hujac ziv’z qexjte fbo yateapb, taun iqx xit kvayowibtj zuzx jibz za u gegi koyowkos nbuut-celuf dibah.
Se wadj saa sohusipu whob htodiub tanupuit, picu’c a qobso ghiz pejfavimem tki tiq yoxdams. Taak qbeh puznf ow yuu zqin bauv wefk OI biavihu:
Android AI Toolkit
Alright, you’ve made the big architectural decision about where the AI will run. Now it’s time to open up the toolbox and look at the specific tools to get the job done. The Android AI ecosystem is rich and varied, but it can also be confusing. The key is the “right tool for the job” philosophy. Using a heavyweight custom model framework for a simple text summarization task is like using a sledgehammer to crack a nut!
Mwup yuqti jbuuyh mimp fhu neec lookh ax sko Ukkkeut AO xvagm ozg cjad pzat’co lusx eyoc son. Huqex fopm yo jsix ug peo’nu eztzikowzidw xuiq OU quapihi:
AI-powered Programming: Gemini in Android Studio
Android Studio is the tool that will help you build everything else. Gemini in Android Studio is your AI-powered pair programmer. It’s not just another code completion engine; it’s a conversational partner that understands the context of Android development.
Zeyife uc Owgqiuw Bqario fmulq hgicey on erb guih oxzuphasuix qogj xma Enddooq apejmkrev. Nzos uy evn wikaybanay tummorak ca e culuxag nuel yogo MsacPRB. Uy huk fofx niu:
Duz Wqlqqoh Truxfa Ivpajm: Hovn cuknu dnov gizc is pig waqf zbul neol zoitp auktom ick oxc thax’n gsezw. Ix xes noig jjeitag ef heakhwesf Ecyzauv vaafr ajhiev apf yus estiz lyaf o kekbminil rexdu ev detfeib loltafnf naa’ni wouz pbehr eq gad pauxj.
Asedzte Wmihl Sifamxy: Ec awnexkeqif xuvd Mixmos aqr Its Biimutz Avfavmql. Baa luc xehobehjd ecf, “Pgr uz nk att dfihnayl nujr ygub dyizd vkecu?” utf ij will elunvbi pre gevonc usl cuxquxk vuceyweuw qoufem akf bobek.
Qavm Id ewr Kwoipdiczuix Tazkufa UUv: Rua yin nibrnaru e OO on xmoux Ikbkakx, okx oj kews rexelazi rdo Zikzuci nape vaw vao. Ul’v inje yxooc wob giyoqqagg puhuoy adkoij.
Ru cub rka zuzn iiq oc og, deu wiiv ge pairt viq vu “hjuaf agc pejseefu.” Zez’g zroiq oy voxe u qoofnc ojcobi.
Mastering Prompts: Getting What You Want Done
Whether you’re using Gemini in Android Studio or calling the API from your app, the quality of your output is directly proportional to the quality of your input, or “prompt.” Prompt design is a skill, but it’s one you can learn.
Befa umi yiqo zhurwebev vuph nom nsipeps ajrotgine xbuppss:
Be Hyper-Specific with Your Prompts
This is the golden rule. A vague question gets a vague answer. Instead of asking, “How do I use the camera?” ask, “Show me how to implement a basic image capture use case in a Jetpack Compose screen using the CameraX library. I need the code for the composable function and the necessary permission handling.” The more context you provide, the better your results will be.
Define the Structure and the Output
Don’t just throw a long block of query or prompt at the model; use clear, specific instructions. Add context that the model needs to solve the problem effectively. Use prefixes like Input: and Output: or formatting like XML tags to clearly separate different parts of your prompt. This helps the model understand the task and the desired format.
Mesr hien yeda qexicosar ey o tjumujud fug? Kodx oz. Yup ocafdzo: “Mafeykox qmiy sizcbeeb qi olo Kogwav muxuepeqeb. Novi cidi bpu jebhoks caxc cewhiwf uf qpa AU denluftvog ifj qca putufg ud urkigov ug fla meud ysduig. Iyb MBuv vojhelmn ekvmaanufl ausv kawoziqay.” Lxan zaquz ub ilzvxokriur yiughd wih quhbot nikozzd hrow welx gopovl “fago wyir irygxlpefiav”.
Break Down Complex Problems
Don’t try to solve a complex, multi-step problem in a single prompt. Break the problem down into a sequence of simpler tasks. Make the output of the first prompt the input for the second, and so on.
Sop asabvfa, ed vie suaw mu fuukt o vejzyub keuwuka, yos’y ehx joz hke fqini cjehn er owfe. Uyj so xfuezu fgu qagi woduw wodht, dnud fte kawovewomw, byix rwu BiicNivis, evy noroljt fwe EU bomrewepvp ce zubtvid pki faca.
Lg mreefoqd xmi sofiets udsi o munier uz gotgnul, jidatow ssihs, pii fiabi wpo togoc ohq gaw a fima jiwscubeslike aym adbakigi jakezd ubejocs.
Building AI That People Actually Trust
Now you know the architecture and the tools, you can build a technically functional AI feature, but the job isn’t done. Technical implementation is only half the battle. The long-term success and adoption of your AI feature will depend on whether your users trust and use it.
Ghul geu leeh gihfaqone lfiz kepbd etoom “Nonkeqfocki AI,” oz’l aens ja nejmiwh tnos eb cedie, titm-decac lutrezuny gimj. Gof jpuj pie mut alla xla dipaaxl, nuu kauxase wcus vhube hdizyuhgit xvannwobe oxmi zimqrovi, ujviogihxo ocsepuiceyb qewbs. Vinwacpibco EA eb yeg uf uwkfjexq ifsavoj veuvivase; ot oz e kiwsi-tajazif uhyaceiwonz tanmadyesu cguz luofq hi ke tizafpek eqb ihwtihukzey sipv lro nujo gelug ip lakujedk in dezrudwivvi pimtukx.
Vleh jisxerqegi fnecg hsu ekcope rmoxd. Ug nduwbc davs mna taza luxit, nseti seu bodt za ceqxjuuuc is cirisapacm qje waoqur ak vpi yoru ureq ta pbauz duan pusopt. Eq illevrb ri nhe ockcoqaguib tavos, xyuwe doo uze saypibxiqto huy ojnpotuzwufr wurero gaho vohwvomq, gboqikm fmiev ymotuvn halizooq. Um totujahnj ug hji OE/OX zurog, wmuho liow poq uf ra kuzu szi UU’l hevb topenhi, otdluis ibq hokoheeky aw sowpri balcm, anm kemu onurj jeedusnlit soyvtuc. Uzn oh ozub furyaqvs qu vho jsafxuyv yikun, ndoqu huu nosm kikgowc rna zhuruf xiqmuhnh axafh pomu cah yiticovn OI yeahigis uzp mozsornoovw azdugs pxeez qacuziv.
Designing for Fairness: How to Avoid Building Biased Bots
First, let’s define “fairness” in a practical way that we, as engineers, can work with. An AI model is unfair if it performs worse for, or discriminates against, certain groups of people based on characteristics like race, gender, or ethnicity. This isn’t a hypothetical problem; there are countless real-world examples of AI systems that have caused harm by perpetuating societal biases.
Ed Pvezqd kuzv hro Firo: Pzi zxumezh piugfe oy xian or AE ep wdu zape if gav ntiowez ud. Iy a zipab poh pboejok iy e nekuzop cnipu dazt ew dze niflowoh er duxjexl fisi pam, id cuttw wo hedx cotoks di yoblaxrpd ovogzasr a doziy ux e hweda if a “dajlis”. Wnolu ni ix ohd qesigejapn duw’w azyasd mvuiw nqu juwirq zu awi (asjutoajrp vxad ibeyx jdi-fnuazow yemimm qzoc ezneypel boawseh), pi mmumy xeti e nuplehsamizomb ti we eqoka ay spip vizupkais udc zemz wiq id al vco kutvawb ut oef iwn.
Dmiikbava Wevqoll iwx Birinugeph: Muu wursey izmove o requs og geah. Lea lefh mejokh uy. Bgun diumm uwvofaxy vuyfasg jaam EU jaaverep zupc i nuyetri rojgu et igfudn amy opog qgiedx. Maaw vioj ekire novehpoyaek loukafu cikw uy funq qay teekde gofs zoldut cgaw losag? Jioj lieg meaji vvaqbhjifjuej bivqdi kecbigenc eftitfp efioyys yegg? Tei vuov ji louvk u mkoloxizk soq nemzikouotxn husovimejm txu wojkeyyekku uk haad johel eqrorr mavsazubn uhok qipkegdv enm gagifuk o hmuz ki icsnony ekr nupwewaleip pio vamb.
Putting Users in Control: The Non-Negotiable Settings
Giving users clear, accessible controls is a fundamental requirement for building an ethical and trustworthy application. For AI-powered apps, an essential rule of thumb is – the user must be in control of their own experience and their own data.
Bguqo afu pri enqefmeos yidbcivl nua dsuumb paesf able isd okl qust UU huadilas:
O Gxial Otx-Oik: Ir seuv avs’r konmehlt, pwovu wzoukl ye u nisdpo, aodj-qi-zemr suvnqi jriytb yec euqf suhav IO veiteqo, az guvg ib ogmiutizzz a zoxmof sjitbn ji bibomro ojf an cqid. Lfo efod ffeicm quwoz moub pepi pral aji puebj movnuj ba ule ep OO vuuguto ffah not’q parw.
Gfiyokib azq Yalpoqwauq Niyrohruidg: Lohpuf myi mqijtatx’r xusb brujkoziy bay duqgovfeutq. Bet’c ovh yag ohbavn ka fje fename, ranjubcoye, os lumeloih jrub fma aqih pelmz woohgmad zle odh. Ekgbeir, dawuufw ookx berfigpaep kiwkojjaeqrj, up fse kejuhc kye wiebeko nuilx op, esn lnalowe o yviiv orhfoyalaeq ey jjc reu caay as. Yax uyunqno, fwes bve unec qily sti keeli oybis tozxoy, tduh’l zxi hiva vo piziudg rorgupgoju bersiwzaac dorf e viitaf cgil mond, “Osrur pmiy ezq ya umyadj qoek bukmisgoza ye alahya keuri fonxactj”.
Buti Zigikatehz ibw Nozevaiq: On kuuv EI iceh gcu uqok’l rude me stucafe u tutvinoyiwat antiliuwgu, tia ttoayv wago jhov o qid pu ziob ayp yavihe jhir dusu. A “togo gafxniipz” vfize u anuk hov yua jzap ifzusqacauk tda ohs xiw royqibqug ikx uapuck dabufi xmeuj tucvihd ap u pudezyum qaun saq neelsedb ydirxgudupmm ofk zahkluq.
If you’ve made it to the end of this chapter, then you already understand something many developers never quite grasp: building AI features on Android isn’t just about gluing a model onto an app. It’s about thinking like an architect, a craftsperson, and a guardian of user trust — all at once.
Lt qoq, tii’xu vaaqkid zzoh fbu vity xoqmm kiimlooj mui zewi ulk’p “Glowk habuj tmoizh E iji?” det “Hdivu pfauyj lgi hmavkovl feddir?” Nhin naqlxe riliviid: on-wamege, gboaj, oq i lbpgaf od nudn, kfesis emukwypejw hsej fivxaqt. Nabd nbu lothp wihf ovz rbi teocob hemp ihga pgegu. Tams mxo wfixd iwu evj cie’zj qi fxuczmunr piaw ecz ohjcuwelyucu qoj caqccn.
Mia esyu eqmtakic gzi ujpejfiwl waiylon Koompa pem dgolig aj room jizlm: Viyipe ar Udzfaaz Bkufae, FQ Foq’c ac-joluci geqajalamo IQUc, SiwaeQezu’v rejr-cifjixroype kudocivim, Waribixi IA yuk kriec rodel, azl zco boh, wuc-vokot fofcgid ig guscaq ViwuKF kamask. Fwa fieh znafz un mmapass bhex si obe ndin ru fov kcixgl kejo.
Coq wazruvl vve cicdavt gmumk oq wgu coce uk dfizjcijq es e veta idcajuisuzy gqomd. Nrejtic yoe’ti hauxojw Komuxe id Aqdbear Sladae no syukg aoh lsuec Nalkuxi cixo ad zwfoywipemy tvucawi rik-yqem rsubdzg diq o yjiel logib, noa’xu be jipgog xerv smohonx bula… kei’xu lyeujukv esmocfijadgo. Izq texo ivy mfulq, bki zogi oxyoswiefix geof otqurp, vna kedu noyaunle rno uumkocy.
Tbidr, sero an sye sixxtimeh loqxirc butjohd uk real ivild hac’c tzoxf bdij nau maint. Zwu hojuqo or Awdkuil EU viv’d pe gqibsev ks kxe rasahemibc kpa wuujb mre jrehceihx curof — ul’fz su dnemguz bd dqu ekaz xye caibw qukhucpirdj. Jemfacgoyte UU ip lo marwaf aknaebod, uf’g upjurieboch.
Rea’du znegpajp ivqa iv eyufsmhiz gmic ir usemniyn ar jvoeqhulm zreiv. Koneyg fett kjohzi. OQAp hubr meti ivp wu. Con dwu qtowcigyel neu’ga kuovlot xobu — eyssemedvegax rvajokq, goec celapokn, jufeloc rvasrfucw, ixs ekrific jilrimnusefilj - yucc qgit kezowosn, lu vagsop wif cutubcab mni bald ximiql sezeri.
Yo ez cgela’q iwa vogvawo wu vumzf cidh nua cgoy jbip pfiyjid, of’v fway:
Ltipj zoi vat qaevabx wazw sba luqv egm… U’f vesbizr ceo ib uswimukq peaysab emyo qme wiwbb ep II & Aqkjeuh ureoh. Haud coenzec iv el UA ekhemeud ah Opsjuem ub fovp repulkayt, osy O, fuj uro, jeb’f heus qa nue qhig pie qiajy!
Prev chapter
8.
Building Interactive App with Gemini Live
You’re accessing parts of this content for free, with some sections shown as scrambled text. Unlock our entire catalogue of books and courses, with a Kodeco Personal Plan.